News How to Set Up PC Case Fans for Airflow and Performance

[Admin]

In this 4Q4RYphfPaSeaZqTePTZpie, we’ll teach you how to arrange the case fans in your PC to unlock maximum performance.

How to Set Up PC Case Fans for Airflow and Performance : Read more

 

[Soaptrail]

In my old case i used the front AND top as intake fans, it worked well since my front of my case had too much blocking the flow of air. As long as the rear exhausts the air it works.

 

[gan14]

I just have to ask. Where did you get these crazy ideas from when you wrote this article.

"Pressure" does not build in computer cases full of wholes. Warmer air also doesn't have time to "travel upwards" in a case where it is ejected in a few milliseconds. If warm air traveled upwards at the speed you think it does, buzzards using updrafts would be ejected into the stratosphere. If you did some tests instead of making MS Paint images, you could have figured those things out by yourself.

 

[TR909]

For the AIO - putting your AIO in pull configuration gets the best cooling for the CPU, in the expense of some higher temps for the GPU - in Push configuration the radiator (being at the top) is fed with hot air coming from CPU, GPU, etc. so not ideal and not really convinced that manufacturers propose this as the 'best case'.

 

[Matt_ogu812]

"...............time you are taking off your side panel to do something inside your system, be sure the system is powered down and unplugged."

So sad to have to tell people common sense things like this.

 

[lmcnabney]

Water cooling systems moving into the case has been a baffling move. When they first gained popularity the water lines would EXIT the case so that the radiator for the CPU's heat did its thing outside of the case. That way the entire case ventilation can be designed to cool the GPU/chipset/RAM without having to worry about the CPU. By having the radiator inside of the case it is either going to be dissipating CPU heat out of the case using pre-heated case air or be more effective cooling the CPU by bringing in external air, but then dumping all of that heat right into the case to make life hard on the GPU. The only good solution is to watercool both the CPU and GPU and blow the radiator(s) out of the case. That way the two big heat generators are having their heat ejected.

 

[wmeyer]

Odd notions in the article with respect to "ideal" configurations. As this is supposed to be a technical discussion, I will assume that all can deal with terms like "eddy currents."
Q: Given a pipe, will it be better to blow air in, or suck it out? A: Assuming the pipe is a simple shape, it won't make much difference. What goes in must come out.
Add some elbows, say three 90 degree turns, and the situation changes. It will be somewhat better to exhaust than to blow in. The reason is that pressurized air hitting the sharp turns will set up eddy currents and these will impede the flow. With an exhaust fan, the air seeks the easiest flow path, and the relative efficiency increases.

Consider then a PC. Multiple intake ports and exhaust ports. Numerous internal obstructions. Exhausting the air will be more efficient. It will also be quieter. Besides the obvious ports, most cases will have leaky gaps through which air may also flow.

A terrible example would be to blow air in through the front of the case where it immediately hits disk drives or other structural obstacles. Those will always create eddy currents.

A simple way to test things is to use smoke. You can purchase smoke tubes, or simply use a lighted cigarette. Moving the smoke source to the vicinity of an intake port you can see how well the air flows. At what distance does the flow cease to deflect the smoke?

Many years ago, Pamotor, a fan manufacturer, offered application notes on how to make cooling most efficient. These days, such references are more difficult to find.

 

[Greg7579]

I just have to ask. Where did you get these crazy ideas from when you wrote this article.

"Pressure" does not build in computer cases full of wholes. Warmer air also doesn't have time to "travel upwards" in a case where it is ejected in a few milliseconds. If warm air traveled upwards at the speed you think it does, buzzards using updrafts would be ejected into the stratosphere. If you did some tests instead of making MS Paint images, you could have figured those things out by yourself.

The author is correct in the way he described this in the very good article. There are nuances and every case is a bit different, but the best way to design airflow in a PC is the way he described it.
I've had this discussion with fellow PC builders several times in the past and there are arguments about how best to design the flow, but in my opinion the author nailed it.
I study high end builds constantly and am amazed at the way some serious enthusiasts place their fans.
The bottom line is if you have a lot of fans in a good case and use an AIO, it is going to cool pretty well even if you screw up the air flow. I just can't imagine why a few of my fellow PC building enthusiasts push hot air into the case through the radiator. I just saw a high-end build that did that....

I think the article is really good and teaches the basics very well. I think he should have mentioned the fan on the PSU because I see a lot of upside-down PSUs for various reasons. In fact, I just built one that way! LOL.

 

[Greg7579]

Odd notions in the article with respect to "ideal" configurations. As this is supposed to be a technical discussion, I will assume that all can deal with terms like "eddy currents."
Q: Given a pipe, will it be better to blow air in, or suck it out? A: Assuming the pipe is a simple shape, it won't make much difference. What goes in must come out.
Add some elbows, say three 90 degree turns, and the situation changes. It will be somewhat better to exhaust than to blow in. The reason is that pressurized air hitting the sharp turns will set up eddy currents and these will impede the flow. With an exhaust fan, the air seeks the easiest flow path, and the relative efficiency increases.

Consider then a PC. Multiple intake ports and exhaust ports. Numerous internal obstructions. Exhausting the air will be more efficient. It will also be quieter. Besides the obvious ports, most cases will have leaky gaps through which air may also flow.

A terrible example would be to blow air in through the front of the case where it immediately hits disk drives or other structural obstacles. Those will always create eddy currents.

A simple way to test things is to use smoke. You can purchase smoke tubes, or simply use a lighted cigarette. Moving the smoke source to the vicinity of an intake port you can see how well the air flows. At what distance does the flow cease to deflect the smoke?

Many years ago, Pamotor, a fan manufacturer, offered application notes on how to make cooling most efficient. These days, such references are more difficult to find.

Disagree on your "blow air through the front" comment. I would probably never exhaust out the front unless I had a radiatir there. I suck in with three big fans in the front and exhaust out the back and through the AIO on the top. I suck in from the front and bottom every time. But I build with expensive cases with good filters. I like pos pressure in the case to force air out through the tiny unfiltered spaces to keep it cleaner. All air coming in via fans is through a filter.

 

[Greg7579]

For the AIO - putting your AIO in pull configuration gets the best cooling for the CPU, in the expense of some higher temps for the GPU - in Push configuration the radiator (being at the top) is fed with hot air coming from CPU, GPU, etc. so not ideal and not really convinced that manufacturers propose this as the 'best case'.

I have had this argument many times and I think you are wrong. I have never understood the argument. Never suck air into the case through a hot radiator. Never.
The author is correct. (Maybe I misunderstood you.)

 

[newtechldtech]

I have had this argument many times and I think you are wrong. I have never understood the argument. Never suck air into the case through a hot radiator. Never.
The author is correct. (Maybe I misunderstood you.)

sadly the author and many of the people never use wind tunnels with AIO in which sucking the air into the case is the best but then wind tunnel them to the near top fans.

Sadly all what people do here is standard Mounting without any ANY innovation.

the best cooling is rad to the front taking in air and then wind tunnel to the nearest top fan front) and and optional rad at the top optional )

 

[Phaaze88]

Yikes. This one is a landmine. I should go get some popcorn...

BTW, chimney config is best config. All these front in, top and rear out discussions over the years, get dunked on!
/S

putting your AIO in pull configuration gets the best cooling

The fans themselves(blade design, size of the central hub, and shape of fan frame) are the reason why that statement isn't true.
Some fans are more effective as push than pull, and vice versa.

I just can't imagine why a few of my fellow PC building enthusiasts push hot air into the case through the radiator. I just saw a high-end build that did that....

It's overblown, when the energy is going to make its way into one's room anyway. [If the PC's in another room, well that's different.]
The overlooked elephant in the room is the gpu, which does the same thing(scattering it's waste heat everywhere), but much higher heat on average when active; power use = heat, not temperature.
If one is browsing on the PC, watching movies, or doing work that uses the cpu... sure, top exhaust the cpu AIO.
Playing games, or gpu-based work... front intake the cpu AIO.
The temperature increase that a front intake cpu AIO does to the components behind it and case ambient pales to what gpus can do to them, especially with this growing adoption of 300w plus models - at least on the Nvidia side so far.
Unfortunate that gpu AIO and custom loop is in a niche spot. Top exhaust the radiator for those, and it does what a top mounted cpu liquid cooler does, but more effective.

But as said initially, it's overblown. Folks should set cooling up how they want to, except when it goes awry.

 

[King_V]

It's overblown

...

But as said initially, it's overblown.

Pun fully intended. Twice. 😆

 

[wmeyer]

Disagree on your "blow air through the front" comment. I would probably never exhaust out the front unless I had a radiatir there. I suck in with three big fans in the front and exhaust out the back and through the AIO on the top. I suck in from the front and bottom every time. But I build with expensive cases with good filters. I like pos pressure in the case to force air out through the tiny unfiltered spaces to keep it cleaner. All air coming in via fans is through a filter.

I was insufficiently clear in that comment. I meant using intake fans in the front. They inevitably lead to eddy currents, and are less effective than they might otherwise be. Exhausting air out the back is generally sufficient. It goes back to the notion of a pipe: What comes out had to have gone in. If you run your PC with the side panel off, and put it on while it is running, and observe a change in the pitch of the fans, then the available intake is inadequate. That is solved by more or larger intake area, not by force.

 

[Greg7579]

Odd notions in the article with respect to "ideal" configurations. As this is supposed to be a technical discussion, I will assume that all can deal with terms like "eddy currents."
Q: Given a pipe, will it be better to blow air in, or suck it out? A: Assuming the pipe is a simple shape, it won't make much difference. What goes in must come out.
Add some elbows, say three 90 degree turns, and the situation changes. It will be somewhat better to exhaust than to blow in. The reason is that pressurized air hitting the sharp turns will set up eddy currents and these will impede the flow. With an exhaust fan, the air seeks the easiest flow path, and the relative efficiency increases.

Consider then a PC. Multiple intake ports and exhaust ports. Numerous internal obstructions. Exhausting the air will be more efficient. It will also be quieter. Besides the obvious ports, most cases will have leaky gaps through which air may also flow.

A terrible example would be to blow air in through the front of the case where it immediately hits disk drives or other structural obstacles. Those will always create eddy currents.

A simple way to test things is to use smoke. You can purchase smoke tubes, or simply use a lighted cigarette. Moving the smoke source to the vicinity of an intake port you can see how well the air flows. At what distance does the flow cease to deflect the smoke?

Many years ago, Pamotor, a fan manufacturer, offered application notes on how to make cooling most efficient. These days, such references are more difficult to find.

I disagree with your take. The author's notions are not odd at all. He is corre

Yikes. This one is a landmine. I should go get some popcorn...

BTW, chimney config is best config. All these front in, top and rear out discussions over the years, get dunked on!
/S


The fans themselves(blade design, size of the central hub, and shape of fan frame) are the reason why that statement isn't true.
Some fans are more effective as push than pull, and vice versa.



It's overblown, when the energy is going to make its way into one's room anyway. [If the PC's in another room, well that's different.]
The overlooked elephant in the room is the gpu, which does the same thing(scattering it's waste heat everywhere), but much higher heat on average when active; power use = heat, not temperature.
If one is browsing on the PC, watching movies, or doing work that uses the cpu... sure, top exhaust the cpu AIO.
Playing games, or gpu-based work... front intake the cpu AIO.
The temperature increase that a front intake cpu AIO does to the components behind it and case ambient pales to what gpus can do to them, especially with this growing adoption of 300w plus models - at least on the Nvidia side so far.
Unfortunate that gpu AIO and custom loop is in a niche spot. Top exhaust the radiator for those, and it does what a top mounted cpu liquid cooler does, but more effective.

But as said initially, it's overblown. Folks should set cooling up how they want to, except when it goes awry.

The back and forth is hard to follow on this interface, which is a bit awkward, especially if on a phone. I hear you and I have had this argument many times for years. I would love to be able to just chat with you about it, but it is difficult here. I'm sure your way will work for a high-end builder and ... you and I are probably very close on temps on our rigs doing it the opposite way. I build in big expensive cases that are very quiet and have great air flow. Show me 9 fans pointed in any direction and the rig is going to cool just fine with an AIO.
But the author taught it the right way, especially for first-time builders. Your way may be OK (but no way would I do it), and maybe I am missing something here, but the idea of sucking cold air through a radiator into the case and making it hot air defies all building logic.... But I hear you and I know there are high-end guys who do it. But I think they are showing off and being contrarian! LOL.
Seriously, don't tell new builders to suck air into a case through a hot radiator, especially from above. Let's just keep it simple.
But things are going to run even hotter now with the latest gen Intel or AMD chips and the new hot GPUs.
But I seriously doubt if I'll being pulling cold air into a hot radiator from above or anywhere else.
Plus, we are going to have to cool the new PCIe Gen 5 SSDs! Ha... And what about the new PSUs to connect and run these new GPUs? Hot.

 

[Phaaze88]

maybe I am missing something here

It's that some folks automatically translate high operating temperature or averages directly to the heat being released. Power use becomes heat, and that heat is going somewhere. Energy changes form - it doesn't just vanish.

A question(s) that should be thrown at said folks is: Which is responsible for heating up your case ambient - and ultimately the room - faster, and makes summer time harder to deal with [assuming constant temperature and power values, which aren't possible in practice]?
Blender + Unigine Heaven/Superposition
Ryzen 7950X at 95C and 235w
RTX 3090Ti at 75C and 500w

Games
12900K at 70C and 90w
RX 6950XT at 80C and 300w

Sitting at the desktop doing much of nothing
12700K at 30C and 55w
RTX 3090 at 30-50C(fan stop on/off) and 15w
[Power values were taken from TechPowerUp reviews and gpu vbios archive. Some rounding was applied.]

If they didn't answer with 2, 2, 1, or 2, 2, 'too low to care'...
¯\(ツ)/¯

Here's another one: If they were using a gpu AIO instead of a cpu one, where would they install it?
At the front? That's really not much different from what an air cooled model does.

Turn it back around to the habit of top exhausting cpu AIOs, and... they're in a similar place as cpu tower air coolers, save for the larger volume of fluid inside.
AIOs tend to have faster and louder stock fans than air coolers, and once the 2 are noise normalized to user comfort(few run their fans at 100%), the gap between them shrinks, especially with the smaller models like the common 240mm.

As long as folks aren't trapping heat inside the PC, I shouldn't care too much...

 

[thestryker]

I think overall this article was pretty good at nailing the basics. What I'd really like to see is a practical test of some of the layouts in something like the Lian-Li o11 EVO where you can mount radiators pretty much anywhere and have airflow through top/bottom/side/front/back.

My old setup was in a Thermaltake F51 where the front is largely closed off so I use an AIO for CPU exhausting through the top. This works really well the vast majority of the time since the CPU doesn't get a heavy load when gaming. Currently I have a 5000D with the CPU AIO exhausting out the side and GPU AIO exhausting out the top. I wish the o11 EVO had been announced a couple weeks or so earlier, because I'd love to have bottom/front intake with top GPU/side CPU.

 

[TR909]

I have had this argument many times and I think you are wrong. I have never understood the argument. Never suck air into the case through a hot radiator. Never.
The author is correct. (Maybe I misunderstood you.)

Does radiator ever gets hot? Try and see how 'hot' it gets.
You could do your research, this is a very old topic (to the point of being trivial) and there are plenty of videos in YouTube (best from Gamers Nexus imho) which are pretty explanatory.

 

[boe rhae]

There's some good info here!

I have two questions if anyone is willing to answer:

I wanted to ask about the idea that pull is less effective than push. I've never heard that before; my understanding is that they're the same or within a margin of error of each other in almost all applications. Is that not correct? Did I misunderstand something?

Also, I'm in the middle of buying a ton of stuff for a custom loop. It's a compatibility nightmare and very expensive, but I thought it'd be cool to say "I built that." I wanted to do a mini-ITX custom loop because SFF builds are so awesome but I didn't trust myself to be patient with something as small and finicky as that, so I went with a Define 7 and got a distro plate for the case. During this process I'd initially decided to do a 360mm radiator setup up top in push/pull, but ended up going with just push. I went with Noctua fans and didn't want to pay another 80 bucks for three more. I haven't bought the CPU or GPU for this build yet (I have a Z690 board but wanted to see if Raptor or Alder would be a better deal before getting a CPU, and I was gonna get a 40xx card from EVGA but... yikes... Probably gonna pick Intel or AMD instead of nVidia for the first time). Do you think it's worth it to go push/pull or should I be fine? I'll be cooling the CPU and GPU. This is a very specific question but if anyone has an answer I'd appreciate it.

 

[Exploding PSU]

This is a very solid article. To be honest, I really never gave the fans much thought. That said, I used to not care about the noise level of the fans, until I installed 6 of them, that's when the realization kicks in.

I'm going to get the boot for asking this, but I have an off topic question : could anyone tell me what is the name of the parts / kits that would allow me to mount the GPU vertically like in the thumbnail image? I've always wanted one, but I didn't even know what it's called.
I reckon it's probably going to need a special case and not just a metal stand and PCIe extender cables like the standard vertical GPU configuration. I would be very grateful if someone can point me to the right direction.

 

[Phaaze88]

Could anyone tell me what is the name of the parts / kits that would allow me to mount the GPU vertically like in the thumbnail image? I've always wanted one, but I didn't even know what it's called.
I reckon it's probably going to need a special case and not just a metal stand and PCIe extender cables like the standard vertical GPU configuration. I would be very grateful if someone can point me to the right direction.

It's this: https://lian-li.com/product/o11de-2/ inside of an O11 Evo(currently using both).

A kit that lets you do that in other cases, I don't know of them.

 

[bit_user]

A positive configuration makes sure your components are always fed with fresh, cool air (which has the side benefit of pushing dust away from your computer).

Wow, a whole article on case fan configuration and only one mention of dust??

And no, positive pressure isn't inherently better at preventing dust buildup, unless you have otherwise open vents at the top of the case. That's the only scenario where I see positive pressure "pushing dust away from your computer", and of course presumes that you don't place intake fans there (like was tried in one absurd test LTT did, in a video I wont' be linking).

The main way positive pressure defends against dust buildup is when used in conjunction with dust filters on the intake fans. Please excuse the bold, but I cannot stress this enough.

Also, I think some actual benchmarks would be nice.

 

[bit_user]

I wanted to ask about the idea that pull is less effective than push. I've never heard that before; my understanding is that they're the same or within a margin of error of each other in almost all applications. Is that not correct?

I did once see an article demonstrating that one positive pressure configuration they tested was much more effective at preventing hotspots, but it was years ago and I can no longer locate it. I don't know if that can be generalized, or if it was specific to that configuration.

I exclusively use positive pressure, but that's really just because I hate dust in my cases. So, I only buy cases with dust filters on all the intake fans, and I make sure the cumulative CFM of the intake fan(s) exceeds that of the exhaust fan(s). Works beautifully!

Also, I use aluminum cases without windows. Aluminum is a good heat conductor (better than steel), while all window materials are poor heat conductors.

Do you think it's worth it to go push/pull or should I be fine? I'll be cooling the CPU and GPU. This is a very specific question but if anyone has an answer I'd appreciate it.

You can probably guess where I stand.
; )

I do only air cooling, but I wonder how well it'd work to use a top-mounted radiator, block off the rear exhaust vents of the case, and just set up intake fans on the front/bottom of the case. I guess whether it'd be enough airflow depends a lot on how much heat you're trying to dissipate, as well as the ambient room temp. I realize it's probably not an option for SFF builds like yours.

 

[Colif]

In this 4Q4RYphfPaSeaZqTePTZpie, we’ll teach you how to arrange the case fans in your PC to unlock maximum performance.

bot posting link exploded it seems

Also note that, when possible, it's best to mount your AIO at the top of your case (or the rear if it's a 120mm AIO) rather than the front. This way, the heat that the AIO moves from your CPU to the radiator gets immediately exhausted out of the case.

oh goody, I just install a 4090 in my case and cook my CPU while I play

isn't it better to have the CPU get fresh air instead of cooking it?

 

[bit_user]

oh goody, I just install a 4090 in my case and cook my CPU while I play

isn't it better to have the CPU get fresh air instead of cooking it?

I guess if you had a rear exhaust fan, that could get rid of a lot of the GPU's heat, leaving the top radiator to draw mostly from the front intake.